Dave Finkelnburg on mon 7 jan 02
Hello all!
Be warned, if you are not a glaze geek you may want to delete now!
Three years ago David Hewitt reported that he and Mike Bailey had done
some glaze testing and observed that additions of cobalt or copper or iron
to a base glaze reduced crazing of the glaze when it was fired to the same
cone on the same clay body.
I have observed that these colorants, when added in significant
quantities, also tend to either flux a glaze or at least change its
viscosity. For example, a common error for some new potter is to add more
than 3% cobalt oxide to a glaze, and then have that glaze, already dark blue
at 1.5% of the oxide, run right off the pot due to the excessive cobalt
addition. :-(
Does anyone have any specific information or even observations which
might answer the question -- is the colorant oxide fluxing the glaze? Is it
simply lowering the viscosity of the melted glaze?
Any technical insight here would be greatly appreciated!
Dave Finkelnburg, obviously avoiding mucking out the glaze room in
Idaho
David Hewitt on thu 10 jan 02
Thank you Dave for raising this question again.
I feel sure that someone must have experienced the effect of adding
colouring oxides to a glaze and seeing the crazing reduced with greater
additions of the colouring oxide.
If you could share this experience with us, we would be most grateful.
It would make a good subject for some properly organised tests - if only
I had the time.
David
In message , Dave Finkelnburg writes
>Hello all!
> Be warned, if you are not a glaze geek you may want to delete now!
> Three years ago David Hewitt reported that he and Mike Bailey had don=
>e
>some glaze testing and observed that additions of cobalt or copper or iro=
>n
>to a base glaze reduced crazing of the glaze when it was fired to the sam=
>e
>cone on the same clay body.
> I have observed that these colorants, when added in significant
>quantities, also tend to either flux a glaze or at least change its
>viscosity. For example, a common error for some new potter is to add mor=
>e
>than 3% cobalt oxide to a glaze, and then have that glaze, already dark b=
>lue
>at 1.5% of the oxide, run right off the pot due to the excessive cobalt
>addition. :-(
> Does anyone have any specific information or even observations which
>might answer the question -- is the colorant oxide fluxing the glaze? Is=
> it
>simply lowering the viscosity of the melted glaze?
> Any technical insight here would be greatly appreciated!
> Dave Finkelnburg, obviously avoiding mucking out the glaze room i=
>n
>Idaho
--
David Hewitt
David Hewitt Pottery ,
7 Fairfield Road, Caerleon, Newport,
South Wales, NP18 3DQ, UK. Tel:- +44 (0) 1633 420647
FAX:- +44 (0) 870 1617274
Web site http://www.dhpot.demon.co.uk
Michael Banks on sat 12 jan 02
A lot of the common colourant oxides we use, act to melt, fluidise, and
decrease the thermal expansion of glaze.
Metal oxides whose stable electron valency is divalent (2+) or monovalent
(1+) at normal pottery firing temperatures, act as fluxes because of their
interaction with silica chains. Colourants whose stable valency is higher,
e.g: chromium sesquioxide (green chromium oxide), -Chromium III oxide (Cr
3+), act as anti-fluxes.
The mechanism is understood to be as follows:
Silicate liquids (ignoring alumina for the moment) are composed of SiO4
tetrahedra linked to each other via two oxygen bonds into linear chains. The
length of these polymerised chains controls the viscosity and
melting/boiling temperature of the liquid. The two oxygen atoms in each
individual SiO4 tetrahedron involved in bonding to the adjacent SiO4
tetrahedron, are called bridging oxygens by geochemists. If divalent or
monovalent cations are present, they can occupy the two unsatisfied oxygen
bonds of each SiO4 tetrahedron, thereby preventing the tetrahedrons bonding
to their neighbours. In geochemical jargon, they increase the number of
non-bridging oxygens (NBO's).
Manganese, cobalt, copper, and other colourants can act as NBO's (and
thereby strong fluxes), but compared to the alkali metals (Li,Na,K,Cs,Rb) or
alkali earths (Mg,Ca,Sr,Ba), the solubility of the transition metals (except
Zn) are limited. So adding more than say 4% copper oxide to a particular
glaze, will result in separation of a black scum of immiscible CuO.
Viscosity is reduced by NBO's relative to their concentration and the length
of the snipped-off chunks of silica chains. For example, potash and soda
are capable of reducing the chains to single tetrahedra (very fluid melts),
but if alumina is present, single large molecules of feldspar (feldspar
quasicrystals) result, with significant increase in viscosity. Colourants
such as copper though, are not affected by alumina (because copper does not
form an aluminosilicate mineral).
The coefficient of thermal expansion (CTE, aka: COE) and the tensile
strength (TS) of a solidified glaze controls the tendency to craze. This is
dependant on the species present in the glass internal structure. Glass is
frozen silicate liquid, and the species entrapped in a pure glass (without
crystals), are those that were present on the original liquid. Pure silica
glass has a low CTE and high TS. Most NBO oxides increase the CTE and reduce
the TS, so act to increase crazing.
Even the low CTE fluxes magnesia, boron and lithium increase the tendency of
silica glass to craze (and reduce hardness) by weakening the glass
structure -the effect of NBO's. But relative to sodium their effect is more
benign in ceramic glazes because fluxes are a necessary evil to induce
silica to melt below 1700 degrees C. Colourant oxides can have a big effect
on CTE (and probably TS), though I have no data on this, other than
observation.
Certain combinations of colourants (e.g: copper, cobalt, chrome, tin,
zirconium and titanium) noticeably reduce crazing in some glazes, though the
effects are complex and likely related to the presence of low CTE molecular
species, rather than the additive effect of the simple measured CTE's of the
individual oxides. I have observed that adding chrome-tin ceramic colours
to a crazing glaze has a big effect on reducing crazing and such suspended
compounds may reduce crazing by more than one mechanism. The opacifier
zircon (zirconium silicate) does this in three possible ways:
(1) the suspended crystals blocking crack propagation
(2) the low CTE of the undissolved zircon crystals
(3) dissolution of a % of the Zr changing the CTE/TS of the glass
It is obvious that crazing of glazes is not entirely predicted from the sums
of the CTE's of the constituent oxides, but by and large, this works fairly
well in practice (except for lithium! But that is Another Story).
Michael Banks,
Nelson,
Middle Earth
----- Original Message ------
Dave Finkelnburg wrote:
> Three years ago David Hewitt reported that he and Mike Bailey had done
> some glaze testing and observed that additions of cobalt or copper or iron
> to a base glaze reduced crazing of the glaze when it was fired to the same
> cone on the same clay body.
> I have observed that these colorants, when added in significant
> quantities, also tend to either flux a glaze or at least change its
> viscosity. For example, a common error for some new potter is to add more
> than 3% cobalt oxide to a glaze, and then have that glaze, already dark
blue
> at 1.5% of the oxide, run right off the pot due to the excessive cobalt
> addition. :-(
> Does anyone have any specific information or even observations which
> might answer the question --
David Hewitt on sun 13 jan 02
Michael,
Thank you for your most informative response to the question of why some
colouring oxides act as fluxes and can reduce crazing. The more people
like you who share their knowledge are much appreciated, even if it
makes me realise how much I don't know.
Just one question. You mention a number of the colouring oxides but not
iron oxide in any of its forms. Could I prevail on you to say how the
various forms of iron oxide might be expected to react in respect of
fluxing and crazing.
David
In message , Michael Banks writes
>A lot of the common colourant oxides we use, act to melt, fluidise, and
>decrease the thermal expansion of glaze.
>
>Metal oxides whose stable electron valency is divalent (2+) or monovalent
>(1+) at normal pottery firing temperatures, act as fluxes because of thei=
>r
>interaction with silica chains. Colourants whose stable valency is highe=
>r,
>e.g: chromium sesquioxide (green chromium oxide), -Chromium III oxide (Cr
>3+), act as anti-fluxes.
>
>The mechanism is understood to be as follows:
>
>Silicate liquids (ignoring alumina for the moment) are composed of SiO4
>tetrahedra linked to each other via two oxygen bonds into linear chains. =
>The
>length of these polymerised chains controls the viscosity and
>melting/boiling temperature of the liquid. The two oxygen atoms in each
>individual SiO4 tetrahedron involved in bonding to the adjacent SiO4
>tetrahedron, are called bridging oxygens by geochemists. If divalent or
>monovalent cations are present, they can occupy the two unsatisfied oxyge=
>n
>bonds of each SiO4 tetrahedron, thereby preventing the tetrahedrons bondi=
>ng
>to their neighbours. In geochemical jargon, they increase the number of
>non-bridging oxygens (NBO's).
>
>Manganese, cobalt, copper, and other colourants can act as NBO's (and
>thereby strong fluxes), but compared to the alkali metals (Li,Na,K,Cs,Rb)=
> or
>alkali earths (Mg,Ca,Sr,Ba), the solubility of the transition metals (exc=
>ept
>Zn) are limited. So adding more than say 4% copper oxide to a particular
>glaze, will result in separation of a black scum of immiscible CuO.
>
>Viscosity is reduced by NBO's relative to their concentration and the len=
>gth
>of the snipped-off chunks of silica chains. For example, potash and soda
>are capable of reducing the chains to single tetrahedra (very fluid melts=
>),
>but if alumina is present, single large molecules of feldspar (feldspar
>quasicrystals) result, with significant increase in viscosity. Colourant=
>s
>such as copper though, are not affected by alumina (because copper does n=
>ot
>form an aluminosilicate mineral).
>
>The coefficient of thermal expansion (CTE, aka: COE) and the tensile
>strength (TS) of a solidified glaze controls the tendency to craze. This =
>is
>dependant on the species present in the glass internal structure. Glass =
>is
>frozen silicate liquid, and the species entrapped in a pure glass (withou=
>t
>crystals), are those that were present on the original liquid. Pure sili=
>ca
>glass has a low CTE and high TS. Most NBO oxides increase the CTE and red=
>uce
>the TS, so act to increase crazing.
>
>Even the low CTE fluxes magnesia, boron and lithium increase the tendency=
> of
>silica glass to craze (and reduce hardness) by weakening the glass
>structure -the effect of NBO's. But relative to sodium their effect is m=
>ore
>benign in ceramic glazes because fluxes are a necessary evil to induce
>silica to melt below 1700 degrees C. Colourant oxides can have a big eff=
>ect
>on CTE (and probably TS), though I have no data on this, other than
>observation.
>
>Certain combinations of colourants (e.g: copper, cobalt, chrome, tin,
>zirconium and titanium) noticeably reduce crazing in some glazes, though =
>the
>effects are complex and likely related to the presence of low CTE molecul=
>ar
>species, rather than the additive effect of the simple measured CTE's of =
>the
>individual oxides. I have observed that adding chrome-tin ceramic colour=
>s
>to a crazing glaze has a big effect on reducing crazing and such suspende=
>d
>compounds may reduce crazing by more than one mechanism. The opacifier
>zircon (zirconium silicate) does this in three possible ways:
>(1) the suspended crystals blocking crack propagation
>(2) the low CTE of the undissolved zircon crystals
>(3) dissolution of a % of the Zr changing the CTE/TS of the glass
>
>It is obvious that crazing of glazes is not entirely predicted from the s=
>ums
>of the CTE's of the constituent oxides, but by and large, this works fair=
>ly
>well in practice (except for lithium! But that is Another Story).
>
>Michael Banks,
>Nelson,
>Middle Earth
>
>
>----- Original Message ------
>Dave Finkelnburg wrote:
>
>> Three years ago David Hewitt reported that he and Mike Bailey had d=
>one
>> some glaze testing and observed that additions of cobalt or copper or i=
>ron
>> to a base glaze reduced crazing of the glaze when it was fired to the s=
>ame
>> cone on the same clay body.
>> I have observed that these colorants, when added in significant
>> quantities, also tend to either flux a glaze or at least change its
>> viscosity. For example, a common error for some new potter is to add m=
>ore
>> than 3% cobalt oxide to a glaze, and then have that glaze, already dark
>blue
>> at 1.5% of the oxide, run right off the pot due to the excessive cobalt
>> addition. :-(
>> Does anyone have any specific information or even observations whic=
>h
>> might answer the question --
--
David Hewitt
David Hewitt Pottery ,
7 Fairfield Road, Caerleon, Newport,
South Wales, NP18 3DQ, UK. Tel:- +44 (0) 1633 420647
FAX:- +44 (0) 870 1617274
Web site http://www.dhpot.demon.co.uk
iandol on sun 13 jan 02
Dear Michael,
First question.
What is an Anti Flux?. Is it something whic robs a flux of it's fluxing =
power? or is it a refractory material? Please enlighten.
You say <composed of SiO4 tetrahedra linked to each other via two oxygen bonds =
into linear chains.>>
Do you mean three dimensional networks?
You say <the two unsatisfied oxygen bonds of each SiO4 tetrahedron, thereby =
preventing the tetrahedrons bonding to their neighbours. In geochemical =
jargon, they increase the number of non-bridging oxygens>>
As I understand it, from reading L & W, Kingery et al. fluxing oxides =
dissociate providing additional oxygen atoms which act as terminations =
where there were previously bridging oxygen atoms, satisfying the =
vertices of the silica tetrahedra which become oxygen deficient due to =
fracture of bridging sites. Then the free cations take up station within =
the spatial structure. It is the distortion and disruption they cause to =
the silicate network which affects the physical properties. Why am I =
wrong? Have I misinterpreted what I have read? If so, where are my =
errors?
I would like to read some of the original papers from Alfred about this =
topic.
Best regards to Middle Earth,
Ivor Lewis. Redhill, South Australia.
Michael Banks on mon 14 jan 02
I'm pleased you were paying attention Ivor :) A few answers:
Hamer, (1st edition 1975, page 10-11) defines antifluxes as the action of
minerals or oxides that hinder fluxing action. While not a truly scientific
term, I think it has some utility in broad descriptions.
I didn't go into the 3D crosslinking of the chains, and random, meandering
chains would have been a better description. I was primarily focussing on
the chain-terminated action of certain cations, on the scale of a couple of
SiO4 tetrahedra.
Regarding the conversion of bridging oxygens into NBO's, no you are not
wrong, but my description (simplified for brevity), simply didn't go into
the nature of the resulting bonds. The fluxing cations (monovalent &
divalent) are bonded via ionic bonds (which are non-directional). It is the
charge value of the cations which neutralises the unsaturated double
electrons which are involved in the (significantly covalent, directional),
Si-O bonds connecting adjacent SiO4 tetrahedra. The charge on the
transition metal ion is crucial. Trivalent or larger charges don't cut it.
As you say, the cations take up station in the voids between the Si-O
chains, but that is the nature of the non-directional ionic bond. The spare
oxygens terminate the directional (50% covalent,) bonds, simply by virtue of
the electronegativity balance, i.e: that they can bond significantly
covalently. The crucial factor in intitiating NBO's remains the charge on
the intruding cation. I suppose the oxide radicals have neccessary parts to
play, but if they're attached to a trivalent cation... they don't. The
misunderstanding hinges on the phrase "occupy the two unsatisfied oxygen
bonds", which I would opine that the essential elements of the bridging bond
(i.e: the shared outer shell electrons on the oxygens) are in fact occupied
by engaged in the ionic bond with the fluxing cation.
Cheers,
Michael,
Nelson,
NZ
----- Original Message -----
iandol wrote:
Dear Michael,
First question.
What is an Anti Flux?. Is it something whic robs a flux of it's fluxing
power? or is it a refractory material? Please enlighten.
You say <SiO4 tetrahedra linked to each other via two oxygen bonds into linear
chains.>>
Do you mean three dimensional networks?
You say <two unsatisfied oxygen bonds of each SiO4 tetrahedron, thereby preventing
the tetrahedrons bonding to their neighbours. In geochemical jargon, they
increase the number of non-bridging oxygens>>
As I understand it, from reading L & W, Kingery et al. fluxing oxides
dissociate providing additional oxygen atoms which act as terminations where
there were previously bridging oxygen atoms, satisfying the vertices of the
silica tetrahedra which become oxygen deficient due to fracture of bridging
sites. Then the free cations take up station within the spatial structure.
It is the distortion and disruption they cause to the silicate network which
affects the physical properties. Why am I wrong? Have I misinterpreted what
I have read? If so, where are my errors?
I would like to read some of the original papers from Alfred about this
topic.
Best regards to Middle Earth,
Ivor Lewis. Redhill, South Australia.
iandol on tue 15 jan 02
Dear Michael,=20
Thanks for confirming my thoughts to that stage.
Another thing crossed my mind, and it may have occurred to other readers =
to ask, "Where do Ionic Radius and Coordination Number come into the =
scheme of things ? " Are these the reasons why Lithium compounds have =
such an impact on cutting the length of that expansion coeff.?
Best regards,
Ivor
Michael Banks on fri 18 jan 02
Yes Ivor, the small ionic radius of the Li+ ion allows it to assume
tetrahedral coordination with oxygen atoms in the lithium alumino-silicate
minerals petalite, spodumene and eucryptite. Compare to the alkali ions in
feldspars (Na+, K+), which are ionically bonded to the nearest 10 oxygens.
Ceramic bodies having the composition of the three lithium minerals (above)
can have zero or significantly negative expansions with rising temperature
(below their respective lattice disruption temperatures). Lithium in glass
has a moderate expansion (between lead and zinc), and not nearly as low as
magnesia, suggesting the lithium ion cannot achieve the compact tetrahedral
coordination in crystal-free glasses.
I'm not sure about this, but I would expect that the shorter the
oxygen-alkai ionic bond, and the lower the coordination number (thus
concentrating the electromagnetic force), the more the bond will resist
expansion by thermal energy.
Michael,
in NZ
----- Original Message -----
Ivor Lewis wrote:
"Where do Ionic Radius and Coordination Number come into the scheme of
things ? " Are these the reasons why Lithium compounds have such an impact
on cutting the length of that expansion coeff.?
| |
|
